A compspec (Composition spec) is a specification and model for comparing things. Compspec IOR is a plugin for extraction of IOR metadata from applications, and packaging in compatibility specification artifacts. This means that we also maintain the compatibility schema here. To learn more:
- Compspec: the Python library that discovers and loads this plugin.
- Compatibility: of container images and applications to a host environment.
- Compspec Go: the Go library that retrieves artifacts and makes graphs for image selection and scheduling.
Install compspec and the plugin here:
pip install compspec
pip install compspec-ior
Then run an extraction with IOR. You can use defaults, or add any parameters to IOR after the plugin name "ior" Here is how to print to the terminal:
compspec extract ior
IOR output
{
"version": "0.0.0",
"kind": "CompatibilitySpec",
"metadata": {
"name": "compat-experiment",
"schemas": {
"io.compspec.ior": "https://raw.githubusercontent.com/compspec/compspec-ior/main/compspec_ior/schema.json"
}
},
"compatibilities": [
{
"name": "io.compspec.ior",
"version": "0.0.0",
"attributes": {
"version": "4.0.0rc1",
"began": "Thu Feb 22 00:36:12 2024",
"machine": "Linux 2b3ee0c4c948",
"finished": "Thu Feb 22 00:36:12 2024",
"command_line": "ior -O summaryFormat=JSON",
"summary.write.operation": "write",
"summary.write.API": "POSIX",
"summary.write.TestID": 0,
"summary.write.ReferenceNumber": 0,
"summary.write.segmentCount": 1,
"summary.write.blockSize": 1048576,
"summary.write.transferSize": 262144,
"summary.write.numTasks": 1,
"summary.write.tasksPerNode": 1,
"summary.write.repetitions": 1,
"summary.write.filePerProc": 0,
"summary.write.reorderTasks": 0,
"summary.write.taskPerNodeOffset": 1,
"summary.write.reorderTasksRandom": 0,
"summary.write.reorderTasksRandomSeed": 0,
"summary.write.bwMaxMIB": 904.92,
"summary.write.bwMinMIB": 904.92,
"summary.write.bwMeanMIB": 904.92,
"summary.write.bwStdMIB": 0.0,
"summary.write.OPsMax": 3619.6798,
"summary.write.OPsMin": 3619.6798,
"summary.write.OPsMean": 3619.6798,
"summary.write.OPsSD": 0.0,
"summary.write.MeanTime": 0.0011,
"summary.write.xsizeMiB": 1.0,
"summary.read.operation": "read",
"summary.read.API": "POSIX",
"summary.read.TestID": 0,
"summary.read.ReferenceNumber": 0,
"summary.read.segmentCount": 1,
"summary.read.blockSize": 1048576,
"summary.read.transferSize": 262144,
"summary.read.numTasks": 1,
"summary.read.tasksPerNode": 1,
"summary.read.repetitions": 1,
"summary.read.filePerProc": 0,
"summary.read.reorderTasks": 0,
"summary.read.taskPerNodeOffset": 1,
"summary.read.reorderTasksRandom": 0,
"summary.read.reorderTasksRandomSeed": 0,
"summary.read.bwMaxMIB": 6615.6215,
"summary.read.bwMinMIB": 6615.6215,
"summary.read.bwMeanMIB": 6615.6215,
"summary.read.bwStdMIB": 0.0,
"summary.read.OPsMax": 26462.4858,
"summary.read.OPsMin": 26462.4858,
"summary.read.OPsMean": 26462.4858,
"summary.read.OPsSD": 0.0,
"summary.read.MeanTime": 0.0002,
"summary.read.xsizeMiB": 1.0,
"test.0.starttime": "Thu Feb 22 00:36:12 2024",
"test.0.capacity": "1.8 TiB",
"test.0.used_capacity": "20.2%",
"test.0.inodes": "116.4 Mi",
"test.0.used_inodes": "5.3%",
"test.0.parameters.testID": 0,
"test.0.parameters.refnum": 0,
"test.0.parameters.api": "POSIX",
"test.0.parameters.platform": "2b3ee0c4c(Linux)",
"test.0.parameters.testFileName": "testFile",
"test.0.parameters.deadlineForStonewall": 0,
"test.0.parameters.stoneWallingWearOut": 0,
"test.0.parameters.maxTimeDuration": 0,
"test.0.parameters.outlierThreshold": 0,
"test.0.parameters.options": "(null)",
"test.0.parameters.dryRun": 0,
"test.0.parameters.nodes": 1,
"test.0.parameters.memoryPerTask": 0,
"test.0.parameters.memoryPerNode": 0,
"test.0.parameters.tasksPerNode": 1,
"test.0.parameters.repetitions": 1,
"test.0.parameters.multiFile": 0,
"test.0.parameters.interTestDelay": 0,
"test.0.parameters.fsync": 0,
"test.0.parameters.fsyncperwrite": 0,
"test.0.parameters.useExistingTestFile": 0,
"test.0.parameters.uniqueDir": 0,
"test.0.parameters.singleXferAttempt": 0,
"test.0.parameters.readFile": 1,
"test.0.parameters.writeFile": 1,
"test.0.parameters.filePerProc": 0,
"test.0.parameters.reorderTasks": 0,
"test.0.parameters.reorderTasksRandom": 0,
"test.0.parameters.reorderTasksRandomSeed": 0,
"test.0.parameters.randomOffset": 0,
"test.0.parameters.checkWrite": 0,
"test.0.parameters.checkRead": 0,
"test.0.parameters.dataPacketType": 0,
"test.0.parameters.keepFile": 0,
"test.0.parameters.keepFileWithError": 0,
"test.0.parameters.warningAsErrors": 0,
"test.0.parameters.verbose": 0,
"test.0.parameters.data packet type": "g",
"test.0.parameters.setTimeStampSignature/incompressibleSeed": 0,
"test.0.parameters.collective": 0,
"test.0.parameters.segmentCount": 1,
"test.0.parameters.transferSize": 262144,
"test.0.parameters.blockSize": 1048576,
"test.0.options.api": "POSIX",
"test.0.options.apiVersion": "",
"test.0.options.test filename": "testFile",
"test.0.options.access": "single-shared-file",
"test.0.options.type": "independent",
"test.0.options.segments": 1,
"test.0.options.ordering in a file": "sequential",
"test.0.options.ordering inter file": "no tasks offsets",
"test.0.options.nodes": 1,
"test.0.options.tasks": 1,
"test.0.options.clients per node": 1,
"test.0.options.repetitions": 1,
"test.0.options.xfersize": "262144 bytes",
"test.0.options.blocksize": "1 MiB",
"test.0.options.aggregate filesize": "1 MiB",
"test.0.results.0.access": "write",
"test.0.results.0.bwMiB": 904.92,
"test.0.results.0.blockKiB": 1024.0,
"test.0.results.0.xferKiB": 256.0,
"test.0.results.0.iops": 3842.6972,
"test.0.results.0.latency": 0.0003,
"test.0.results.0.openTime": 0.0001,
"test.0.results.0.wrRdTime": 0.001,
"test.0.results.0.closeTime": 0.0,
"test.0.results.0.totalTime": 0.0011,
"test.0.results.1.access": "read",
"test.0.results.1.bwMiB": 6615.6215,
"test.0.results.1.blockKiB": 1024.0,
"test.0.results.1.xferKiB": 256.0,
"test.0.results.1.iops": 27962.0267,
"test.0.results.1.latency": 0.0,
"test.0.results.1.openTime": 0.0,
"test.0.results.1.wrRdTime": 0.0001,
"test.0.results.1.closeTime": 0.0,
"test.0.results.1.totalTime": 0.0002
}
}
]
}
And how to save to file
compspec extract --outfile ior-test.json ior
And run the example to see how to use compspec-ior directly in Python to generate the same artifact.
python ./examples/singleton-run.py
If you open the Development container in VSCode, you'll find ior on the path:
$ which ior
/usr/bin/ior
This allows us to easily develop and test the compatibility plugin. You can
- how to handle adding lists (with indices) to schema?
HPCIC DevTools is distributed under the terms of the MIT license. All new contributions must be made under this license.
See LICENSE, COPYRIGHT, and NOTICE for details.
SPDX-License-Identifier: (MIT)
LLNL-CODE- 842614